Performance comparison of multicast MANET CRN system’s based on POS scheme with different types of router protocol algorithms

0.78 MB·English

by  Basma Nazar Nadhim, Sarab Kamal Mahmood

Indonesian Journal of Electrical Engineering and Computer Science
Vol. 17, No. 1, January 2020, pp. 291~302
ISSN: 2502-4752, DOI: 10.11591/ijeecs.v17.i1.pp291-302  291

Performance comparison of multicast MANET CRN system’s
based on POS scheme with different types of router protocol
algorithms

Basma Nazar Nadhim, Sarab Kamal Mahmood
Electrical Engineering Department, College of engineering, Mustansiriyah University, Iraq

Article Info ABSTRACT
A multi-layer multi-hop mechanism, with one of three different types of
Article history:
router protocol algorithms (Steiner minimal tree (SMT), shortest path tree
Received May 30, 2019 (SPT) and minimal spanning tree (MST)) was employed for the construction
of multi-cast MANET CRNs. The probability of success (POS) used as the
Revised Jun 1, 2019
channel assignment scheme, is dependent on the availability of the channel,
Accepted Jun 15, 2019
and the required transmission time. It was applied to the network after the

conversion of the network‟s random topology into one of the three
algorithms (SMT, SPT or MST). This facilitated the selection of an efficient
Keywords:
channel for the transmission of the CR user‟s data, under the effect of the
CRN Rayleigh fading channel. Three idle probability circumstances (i.e., =0.1,

MANET 0.5, 0.9), and different network parameters, were used to compare the
MST performance of the CRN with three routing protocols in terms of throughput,
and packet delivery rate (PDR). According to the simulation results, for a
SMT
high traffic load of PUs ( 1) at different network parameters, the CRN
SPT
with the SPT algorithm performed better than the CRN with the SMT or
MST algorithm.
Copyright © 2020 Institute of Advanced Engineering and Science.
All rights reserved.
Corresponding Author:
Basma Nazar Nadhim,
Electrical Engineering Department,
Mustansiriyah University, Baghdad, Iraq.
Email: [e-mail]

1. INTRODUCTION
In the last decade, wireless services have gained rapid popularity. Numerous countries have adopted
the fixed spectrum allocation methodology, in which a majority of the available radio spectrum has already
been assigned for different services [1-3]. However, a major chunk of the assigned spectrum has been
employed sporadically and the range of geographical variations in terms of the usage of assigned spectrum
falls between 15% and 85% accompanied by a high variance in time [4]. In fact, as per a recent study done
by the Federal Communications Commission (FCC), majority of such licensed spectrums were still
unoccupied for large time periods [5, 6]. We have put forward dynamic spectrum access (DSA), also referred
to as cognitive radio, as a substitute policy to allow efficient use of the radio spectrum [7, 8]. Thus, it can be
considered as a potential technique that can be applied in future wireless communications to address the
spectrum scarcity issue [9]. As per the FCC, cognitive radios (CRs) are radio systems that can allow
conducting spectrum sensing uninterruptedly, dynamically detect spectrums that are unused and then
functions in those spectrum holes in which there are idle licensed (primary) radio systems [10]. In other
words, in terms of CRN, secondary users (SUs) could employ spectrum access opportunities to perform
unlicensed transmissions when the licensed spectrum is not occupied by primary users (PUs) [11-13]. In Next
Generation (xG) networks, employing cognitive radios can help identify unutilised spectrum as well as
spectrum sharing with no disadvantaged interjecting with other users (Spectrum sensing), capturing the best
available spectrum that is in par with the communication demands of the user (Spectrum management),
Journal homepage: http://ijeecs.iaescore.com
2 9 2  ISSN: 2502-4752
maintain tractable communication exigencies when shifting to better spectrum (Spectrum mobility) and offer
an equitably spectrum that allow assigning method amongst cohabitation xG users (Spectrum sharing )
[4, 14, 15]. Based on this definition, there are two key characteristics pertaining to the cognitive radio:
Cognitive capability that allows information sensing for the radio technology from its radio environment,
and reconfigurability that allows dynamic programming of the radio based on the radio environment [4, 14].
Figure 1 presents the needed mission pertaining to adaptive operation for open spectrum, which is also
referred as the cognitive cycle. The key steps in the cognitive cycle include spectrum analysis, spectrum
sensing and spectrum decision [14]. Numerous key applications are associated with cognitive radio. For
instance, for radio interoperability, cognitive radio plays a key technology in the US military (i.e. JTRS
program), future mobile base stations and public safety (i.e. SAFECOM program). For multi-hop ad hoc
networks, CR is regarded as the radio platform. Multicast is a key service that the ad hoc networks need to
support [16]. A mobile ad-hoc network (MANET) includes mobile nodes with no requirement for
infrastructure [17]. MANET can be considered as a self-organising network without infrastructure, wherein
each of the participating devices can receive as well as send data along with mobility model that is
independent. The MANET ad hoc network has the ability to change locations as well as configure itself [18].
Multicasting is employed when there is a need for applications to send the same data multiple destinations.
The communication costs for applications can be decreased with multicasting, when there is a need to send
the same data to multiple recipients. In place of sending through multiple unicast, multicasting helps to
reduce the link bandwidth consumption, delivery delay and router processing [19]. Classification of the
existing multicast routing protocols pertaining to the MANETs could be done into mesh-based and tree-
based. The difference between these protocols lies based on the path redundancy between receivers and
senders. Tree-based protocols offer just a single path between receivers and senders, while mesh-based
protocols offer multiple paths [20, 21]. In recent years, there have been numerous routing protocols
pertaining to CRNs that have been put forward and examined. More commonly, these protocols focus on
either picking the channel that possesses the maximum average spectrum-availability time or the best-quality
channel. In a CRN, both the needed transmission time and spectrum-availability time can considerably affect
routing and network connectivity. Specifically, a significant reduction in CRN performance spectrum-
availability time could result when there is smaller average spectrum-availability time of an assigned channel
than what is needed for the transmission time over that channel. Even worse, this issue becomes even more
vital in multi-hop CRNs when there are a lot of multiple links involved. Network performance can be
enhanced by employing a diverse channel quality as well as spectrum availability more efficiently, provided
these are considered by the cognitive routing protocol design [22]. Apart from employing efficient channel
assignment to select the best channel for data transmission in multi hop multicast cognitive radio network,
the manner in which network topology‟s connection also plays an important role in enhancing the system
performance. The function of tree-based multicasting protocols is based on the tree construction for the
overall graph that connects all the multicast groups in an acyclic subgraph together. In a tree structure,
through a single path, every node could reach out to any of the other nodes. Two fundamentally separate
approaches can be considered for the construction of multicasting trees: employing minimum cost trees
(MCTs) or „shortest path trees‟ (SPTs). The former approach is aimed at decreasing the overall edge cost of
the tree, while the latter approach helps to reduce each receiver‟s distance from the sender [23].
In this research paper, we have employed the multi-layer multi hop multicast MANET CRN that
included three types of router protocol algorithms (shortest path tree (SPT), Steiner minimal tree (SMT) and
minimal spanning tree (MST)) for CRN. To enhance the network performance with regards to the throughput
and packet delivery rate (PDR), we employed the probability of success (POS) as the channel assignment
scheme, which is applied to the network post transformation of the network‟s random topology to one of the
three router protocol algorithms. This allows selecting an efficient channel for data transmission based on the
channel‟s availability and the needed transmission time. A comparison of the multi-layer multi-hop multicast
MANET CRN system‟s performance along with these three algorithms as well as POS scheme was done to
check if each of the router protocol algorithms shows the best performance versus others at various traffic
loads of PU and different network parameters.
The rest of the paper is structured as follows: Section 2 presents the related work of practices and
methods for multicasting network channel assignment and the routing protocol for undirected graph.
Section 3 presents the router protocol algorithms for a multicasting network. The system model for the
recommended multicast protocol is presented in Section 4. Lastly, the simulation outcomes and conclusions
are presented in Sections 5 and 6, respectively.

Indonesian J Elec Eng & Comp Sci, Vol. 17, No. 1, January 2020 : 291 - 302
Indonesian J Elec Eng & Comp Sci ISSN: 2502-4752  293

Figure1. Cognitive radio cycle

2. RELATED WORKS
In cognitive radio networks, multicasting seems a challenging issue because of the dynamic nature
pertaining to the available spectrum opportunities for secondary users. The multicast cognitive radio
network‟s performance is improved with regards to the throughput and various parameters by employing
different methods and techniques for channel assignment as well as constructing the routing tree from
random undirected topology [23]. To establish efficiency, there is a need for tree-shaped topologies
pertaining to multicast connections. Transmission of a minimum number of data packets could be done in
parallel to several destinations along the tree branches, where duplication is performed only for tree branches
[24]. In [24], identification of Steiner trees pertaining to the topology of multicast connections for
communication networks can be done when the primary aim is cost optimisation. In [25], presentation of the
method pertaining to transforming undirected topology to Steiner tree was done. In [26], a novel routing
protocol was put forward for multichannel CRNs, which uses a probabilistic metric. This approach relies on
probabilistically determining the available capacity pertaining to each channel over every CR-to-CR link,
while also considering primary radio (PR). In [27], CoCast was put forward as an ad hoc multicast protocol
pertaining to cognitive radio-based MANETs. CoCast mitigates the scalability issue of ODMRP with regards
to the number of multicast sources that use multiple cognitive radio channels. For building the minimum-
energy multicast tree, a low complexity approximation algorithm guarantees bounded performance, which
converts the multicast issue to a directed Steiner tree issue as mentioned in [28]. In terms of the cognitive
radio networks, this approximation algorithm considers the energy that was utilised for sensing the spectrum
opportunities, while its constructed multicast trees were found to be adaptive towards the traffic load
pertaining to the primary network. In [29], a cross-layer optimisation approach has been put forward for the
multicast video in CR networks. The modelling of the CR video multicast is done as an optimisation problem
by accounting for important design factors such as video rate control, scalable video coding, spectrum
sensing, modulation, scheduling, dynamic spectrum access, primary user protection and retransmission.
For a multiuser single-transceiver cognitive radio network (CRN), the coordinated spectrum access issue has
been regarded in [30]. Our goal here is to increase the sum-rate that has been realised through all contending
cognitive radio users with regards to both transmission rate and spectrum assignment. In cognitive radio
mesh networks, the multicast routing as well as channel allocation problem was accounted in [31].
An algorithm was put forward that concierges switching latency and channel heterogeneity. The algorithm
focuses on decreasing the end-to-end delay, and simultaneously minimising the degradation of throughput by
employing a dynamic programming approach. In [22], a novel routing metric was put forward for multi-hop
CRN that falls under Rayleigh fading channel. This routing metric tries to maximise the success probability
pertaining to a given CR transmission by accounting for both required transmission times as well as average
residual spectrum-availability. With regards to this metric, a routing protocol was built for multi-hop CRN,
namely MaxPoS. In [32], for the throughput maximisation in cognitive radio networks, overlapping and non-
overlapping channel assignment algorithms were accounted.
In [33], a routing algorithm has been put forward that makes use of the expected transmission count
metric (ETX) pertaining to multi hop wireless cognitive network to choose high-quality channels to perform
routing on a hop-by-hop basis. In [34], a cross-layer multicasting routing protocol has been put forward to
stream video over cognitive radio networks to improve the received video‟s overall quality in the multicast
group. In [35] and [36], MST and SPT are employed as routing protocols pertaining to multilayer multicast
multi hop CRN along with POS scheme employed as channel assignment. The network‟s performance is
Performance comparison of multicast MANET CRN system’s based on POS… (Basma Nazar Nadhim)
2 9 4  ISSN: 2502-4752
improved with regards to throughput and PDR versus employing other schemes. As routing algorithms,
EXT with SPT and MST are used for multicast multi hop while the POS scheme was employed as channel
assignment to select an efficient channel for data transmission based on the channel availability and the
needed time for transmission as presented in [37].

3. ROUTER PROTOCOL ALGORITHMS FOR WIRELESS MULTICASTING
The analysis and design of wireless multicast employs different algorithms and techniques. In this
section, three types of router protocol algorithms (minimal spanning tree (MST), shortest path tree (SPT) and
Steiner minimal tree (SMT)) have been introduced as mentioned below:

3.1. Shortest Path Tree (SPT) Algorithm
The SPT is a kind of tree created from an undirected graph. An undirected graph has weights that
are non-negative and a node specially designated as root or source. The shortest path tree is the path having
minimum total weight starting from the source and going towards each of the destinations in the graph.
In case there is a change in the source node, the SPT should be re-created, making it more complex.
The shortest path tree can be created with Dijkstra‟s algorithm having an overall running time which is equal
to (m + n log n), where m represents the amount of links (edges) that join the nodes while n denotes the
number of vertices. Figure 3 shows two SPTs having node (a) as source node for the weighted graph shows
in Figure 2. The total weight of the edges in both the SPTs cannot be same. In Figure 3a it is 18, whereas it is
17 in Figure 3b. The purpose is to obtain the group of edges joining all nodes where the sum of the weights
of the edges is minimum from the root to every node [38, 39].

Figure 2. A weighted graph Figure 3. A shortest-paths tree rooted at vertex a

3.2. Minimal Spanning Tree (MST) Algorithm
The minimal spanning tree can be obtained by covering all the nodes in the undirected graph.
It obtains a path having minimum sum of all the edges‟ total weights for every destination, and does not
contain any loop. Many distinct MSTs are possible, for a single undirected graph having same total weight.
Moreover, there is no need of a starting node (that is, the source). MST is flexible where any node can be
selected as a source. Thus, if there is a change in the source node, then the tree need not be re-created.
A minimal spanning tree can be created using Kruskal‟s technique which has a total running time equivalent
to (m log m). Figure 4 shows the MST for the graph in Figure 1. The total weight of the edges in this MST is
14. It is noteworthy that in case of MST, the root (source) node always requires more hops to cover all the
nodes in comparison to SPT for the same graph. In Figure 4, the source node requires 4 hops to get to the last
vertex in the MST, whereas in SPT, the source node requires just 3 hops to get to the last node as given in the
Figure 3 [35, 38, 39].

Figure 4. A minimal spanning tree
Indonesian J Elec Eng & Comp Sci, Vol. 17, No. 1, January 2020 : 291 - 302
Indonesian J Elec Eng & Comp Sci ISSN: 2502-4752  295

3.3. Steiner Minimal Tree (SMT) Algorithm
A Steiner minimal tree spans a specific subset of nodes as against a spanning tree where all the
nodes are spanned. For SMT, the nodes are distributed in two sets: nonterminal and terminal nodes.
The terminal nodes are those nodes which are to be included in the SMT solution. The SMT cost is equal to
the total weight of the edges. A Steiner tree possibly has certain nonterminal nodes to minimise the cost.
Let V represent the set of nodes. As a rule, there is a terminal set L V and a metric identifying any 2
vertices‟ distance from each other in V. The purpose is to obtain a connected subgraph that spans every one
of the terminals having minimum total cost. A renowned technique to obtain an SMT is by using an MST
(minimal spanning tree). First, we build the metric closure on L, i.e. an entire graph with vertices L and edge
weights equalling the shortest path lengths. Then, we determine an MST on the closure, wherein every edge
matches one shortest path on the actual graph. Lastly, the MST is transmuted back to a Steiner tree by
supplanting every edge with the shortest path and certain direct post processing for eliminating any likely
cycle. Figure 5a depicts an undirected graph G with terminal vertices L = {v1, v2, v3, v4, v5} and
nonterminal vertices {u1, u2, u3, u4}. From Figure 5c and Figure 5d, it can be said that prior to adding
{u1, u2}, the tree‟s total weight is 22, whereas post adding {u1, u2}, it is reduced to 17. This means the
Steiner points (nonterminal vertices) play a role in decreasing the total cost of the tree [38, 39].

Figure 5. A Steiner minimal tree

4. SYSTEM MODEL FOR MULTICAST MANET CRN
This paper study considers multilayer multicast multi-hop MANET CRN with three kinds of router
protocol algorithms (Steiner minimal tree (SMT), shortest path tree (SPT) and minimal spanning tree (MST))
for video transmission from the source to destination nodes over a solo session. First, an undirected graph
with set of N vertices is produced within a square area. The algorithms outlined in section (3) are deployed
for constructing MANET CRN. Several PU channels (M) are obtainable between the source and every
destination node. The status model of every primary user (PU) channel is the Markov model, which switches
between two states (idle and busy). A busy state indicates that the channel cannot be utilised by SU, while an
idle state means that the channel is not utilised by PU. For all channels, the bandwidth (BW) set is the same.
The POS scheme is deployed on the network as channel assignment for improving the network performance.
For organising the CRN transmissions, a common control channel (CCC) is presented [35, 37]. The close
form term for probability of success ( ) between any two nodes i and j in multilayer MANET CRN

over channel j which contains the available channel (C) of CRN is expressed in [22] as in (1):

( )
(1)

where is the average availability time of spectrum in (in sec) for channel j and is the requisite

transmission time in (in sec/packet) for sending a packet from node i to k over channel j where it can be
stated as in (2) [22]:
Performance comparison of multicast MANET CRN system’s based on POS… (Basma Nazar Nadhim)
2 9 6  ISSN: 2502-4752
(2)

where D is the packet size (in bits) and is the data rate (in bit/sec) between nodes i and k over channel

j which can be stated as in (3) [22]:

( ) (3)

where signifies the thermal power density in (Watt/Hz), is the channel bandwidth and signifies

the received power from transmitter i to receiver j which can be stated as in (4) [22]:

( ) (4)

where is CR‟s transmission power, d is the distance between any two nodes, n is the path loss exponent,

and is the channel power gain between nodes i and k over channel j. For Rayleigh fading, is

exponentially distributed with mean 1[22].

5. COMPUTER SIMULATION AND RESULTS
Computer simulations for multicast MANET CRN with multi-layer multi hop were performed.
Three different routing protocol algorithms (SMT, SPT and MST), together with a probability of success
(POS) channel assignment scheme, were used to assess the network performance in terms of throughput and
packet delivery rate (PDR), under the effect of the Rayleigh fading channel. Three values for idle probability
(i.e. = 0.1, 0.5 and 0.9) and the various network parameters used for comparison purposes, are displayed in

Table 1. MATLAB R2018a was employed to carry out the simulations.

Table 1. System Parameters
Parameter Value/Type
Network area 200m*200m
No. of Nodes 30
No. of Terminal nodes (Nt) 20
No. of Nonterminal nodes (Nnt) 10
Topology tree (SMT, SPT and MST)
No. of CR source One source
No of primary channel (M) 15
PU channel model Markov model
Idle probability [0.1 0.5 0.9]

Average availability time ( Range from 2ms to 45ms

Bandwidth (BW) 1MHz
Packet size (D) 4KB
Transmission power (Pt) 0.1Watt
Channel used Rayleigh fading channel
Path loss exponent (n) 4
Thermal noise power ( ) W/Hz

5.1. Performance Evaluation of Multicast MANET CRN Under The Impact of Channel Bandwidth
The throughput and PDR performance of multilayer multi hop multicast MANET CRN,
with regards to the channel bandwidth with three different types of tree algorithms (SMT, SPT and MST),
and three idle probability values are presented in Figures 6 and 7 respectively. POS

was used as the channel assignment scheme in all the CNRs. According to the results attained, the data rate is
proportional to the bandwidth of the channel. This indicates that an increase in the channel‟s bandwidth,
improved the performance of the CRN tree algorithms. It was also observed that in terms of throughput and
PDR, an increase in the value enhanced the performance of the CRN at the three algorithm protocols.

This is because a high value raises the probability, that suitable channels will be available for transmission

by CR users at a low traffic load of PU t high and moderate idle probability values

it was observed from Figure 6 that the throughput performance of the CNR using the SMT routing tree
Indonesian J Elec Eng & Comp Sci, Vol. 17, No. 1, January 2020 : 291 - 302
Indonesian J Elec Eng & Comp Sci ISSN: 2502-4752  297

algorithm, is comparable to the performance of the CRN with the MST algorithm at 15.5% and 15.49%,
while outperforming the CRN with the SPT algorithm by 24% and 26.3% at respectively.

At busy traffic of the channel by primary users, where the CRN with the SPT algorithm

outperformed the CRN with SMT and MST algorithms by 46% and 88.72% respectively. As for

, the PDR performance of the CNR using the SMT tree, is comparable to the CRN with the MST
tree at 11.2% and 16.7% respectively, while outperforming the CRN with the SPT algorithm by 16.7% and
24.13% respectively. And finally, at a busy channel ( ), the PDR performance of CRN with SPT

outperformed the CRN with the SMT and MST algorithms by 64.83% and 131.92% respectively (Figure 2).

Figure 6. Throughput vs. channel bandwidth Figure 7. PDR vs. channel bandwidth

5.2. Performance Evaluation of Multicast MANET CRN under the Impact of Packet Size
The throughput and PDR performance of multilayer multi hop multicast MANET CRN,
with regards to the packet size with three different types of tree algorithms (SMT, SPT and MST), and three
values of idle probability are portrayed in Figures 8 and 9 respectively. The channel

assignment scheme used for the CRNs is the POS scheme. According to the figures presented,
the performance of all CRN tree algorithms, in terms of throughput and PDR, declined as the value of packet
size D increased. This decline in performance is due to the fact that at high packet size values, the successful
delivery of data from a CR source to destination nodes, calls for an increased channel availability time.
This renders the search for the best channel difficult. However, as mentioned in Section (5.1), an increase in
the value of enhanced the performance of the CRN at the three algorithm protocols. This enhancement

came about because the traffic load of PU is low at a high value of As dispayed in Figure 8, at idle

probability , the throughput performance of the CRN using the SMT routing tree algorithm,

outperformed the CRN with SPT and MST algorithms by 7.81% and 13.88% at P = 0.9, and by 8.84% and
I
13% at P = 0.5 respectively. However, at the CRN with the SPT algorithm outperformed the CRN
I
with the SMT or MST algorithms by 31% and 46.94% respectively. As shown in Figure 9, at idle probability
the PDR performance of the CNR using the SMT tree routing algorithm is superior to the

CRN with SPT and MST algorithms, and is comparable to the MST algorithms at 16.5% and 8.96% at
0.9, and by 16.16% and 6.1% at respectively. However, at P = 0.1, the CRN with the SPT
I
algorithm outperformed the CRN with the SMT and MST algorithms by 46.17% and 63.31% respectively.

Figure 8. Throughput vs. data packet size Figure 9. PDR vs. data packet size
Performance comparison of multicast MANET CRN system’s based on POS… (Basma Nazar Nadhim)
2 9 8  ISSN: 2502-4752
5.3. Performance Evaluation of Multicast MANET CRN under the Impact of Increased Primary
Channels
The throughput and PDR performance of the multilayer multi hop multicast MANET CRN,
with regards to the number of primary channels with three different types of tree algorithms (SMT, SPT and
MST), and three values of idle probability , are presented in Figures 10 and 11

respectively. The channel assignment scheme used for the CRNs is the POS scheme. The figures revealed
that the throughput and PDR performances were enhanced by an increase in the number of primary channels.
This is because an increase in the number of primary channels creates the opening, for an enhancement in the
number of channels available to CR users. As illustrated in Figure 10, for and number of channels

6, the CRN with the SPT algorithm offers a better performance than the CRN with the SMT and MST
algorithms. Also, the CRN with the SPT algorithm outperforms the CRN with the SMT and MST algorithms
by 85.4% and 135.7%, and by 36.26% and 48.36% at respectively. This is attributed to the

SPT algorithm using the short path distance between source node and destinations. With the use of the short
path distance, the number of hops is minimized to reduce the required transmission time, and increase the
value of throughput. As shown in Figure 11, at and for number of channels , in terms of PDR,

the SPT algorithm is superior to the SMT and MST algorithms. At the performances of the SMT

and MST algorithms are comparable, while outperforming the SPT algorithm. At , the SPT

algorithm outperformed the SMT and MST algorithms by 191.89% and 328.88% respectively.

Figure 10. Throughput vs. Number of channels Figure 11. PDR vs. Number of Channels

5.4. Performance Evaluation of Multicast MANET CRN under the Impact of Increased Transmission
Power
The throughput and PDR performance of multilayer multi hop multicast MANET CRN,
with regards to transmission power with three different types of tree algorithms (SMT, SPT and MST),
and three values of idle probability , are presented in Figures 12 and 13 respectively.

POS was used as the channel assignment scheme for all the CNRs. As the transmission power increased,
the performance of the CRN for the three algorithms were enhanced, in terms of throughput and PDR. This is
attributed to the high value of transmission power reducing the time required for data transmission. As such,
more data can be transmitted over each channel. Also, in a circumstance where the traffic load for PU is low,
the CRN performance for the three algorithms is enhanced (as mentioned in previous sections). The enhanced
throughput gains for the CRN with the SPT algorithm, against the CRN with the SMT and MST algorithms
at , 0.5 and 0.9 are 74.82% and 108.4%, 39% and 61.39%, and 38.93% and 57.42% respectively.

The improved PDR gains for the SPT algorithm in comparison to the SMT and MST algorithms at ,

0.5 and 0.9 are 78.64% and 111.38%, 11.32% and 14.24%, and 8.79% and 13.59% respectively.

Indonesian J Elec Eng & Comp Sci, Vol. 17, No. 1, January 2020 : 291 - 302
Indonesian J Elec Eng & Comp Sci ISSN: 2502-4752  299

Figure 12. Throughput vs. Transmission Power Figure 13. PDR vs. Transmission Power

5.5. Performance Evaluation of Multicast MANET CRN under the Impact of Path Loss Exponent
The throughput and PDR performance of multilayer multi hop multicast MANET CRN with regards
to path loss exponent, with three different types of tree algorithms (SMT, SPT and MST), and three values of
idle probability , are presented in Figures 14 and 15 respectively. POS was used as the

channel assignment scheme for all the CNRs. As the value of path loss exponent (n) increased, the CRN
performance for the three algorithms dipped in terms of throughput and PDR. This dip, which is attributed to
the negative influence of the Rayleigh fading channel, greatly reduced the chances of finding the best channel
for data transmission. Figures 9 and 10 indicate that at n 4, and as the value of is increased,

the performance of the CRN at the three algorithm protocols was improved, and the performance of the CRN
with the SPT algorithm, grew superior to those of the CRN with the SMT or MST algorithm by 172.08% and
247.81%, 114.89% and 166.3%, and 109.65% and 160.03% at idle probability = 0.1, 0.5 and 0.9 respectively
in terms of throughput, and by 82.52% and 117.53%, 13.55% and 20.25%, and 13.17% and 17.97% at idle
probability = 0.1, 0.5 and 0.9 respectively in terms of PDR. This is attributed to the SPT algorithm‟s use of
the short path distance between the source node, and the destinations. The resulting minimized number of
hops reduced the required transmission time, to consequently improve the performance of the CRN with the
SPT algorithm.

Figure 14. Throughput vs. Path loss exponent Figure 15. PDR vs. Path loss exponent

5.6. Performance Evaluation of Multicast MANET CRN under the Impact of Idle Probability
The throughput and PDR performance of the multilayer multi hop multicast MANET CRN, with
regards to the values of idle probability, with three different types of tree algorithms (SMT, SPT and MST),
and with POS as the channel assignment scheme in all the CNRs, are presented in Figures 16 and 17
respectively. As indicated in Figure 11, at idle probability 0.3, the performance of the CRN with the SPT
algorithm during a high traffic load of PUs, is superior to the CRNs with the SMT or MST algorithm. As the
idle probability increased, the performance of all the CRNs improved. This is attributed to the raised
probability, that suitable channels will become available at a low traffic load of PUs. For Idle probability
0.3, the performance of the SMT algorithm is comparable to those of the SPT and MST algorithms at 4.59%
and 11.4% respectively. As portrayed in Figure 17, at idle probability 3, the CRN with the SPT algorithm,
experienced a bigger drop in PDR performance than the CRNs with the SMT or MST algorithm.
Performance comparison of multicast MANET CRN system’s based on POS… (Basma Nazar Nadhim)
3 0 0  ISSN: 2502-4752

Figure 16. Throughput vs. Idle probability Figure 17. PDR vs. Idle probability

6. CONCLUSION
During this investigation, we delved into the use of a multi-layer multi-hop mechanism, with one of
three different types of router protocol algorithms (Steiner minimal tree (SMT), shortest path tree (SPT) and
minimal spanning tree (MST)), for the construction of multi-cast CRNs. The selection of an efficient channel,
for data transmission by CR users, was conducted under the effect of the Rayleigh fading channel. The
probability of success (POS) was employed as the channel assignment scheme. Three values for idle
probability (i.e. =0.1, 0.5 and 0.9), and different network parameters, were used to compare the CRN

performance with regard to three routing protocols, in terms of throughput and PDR. According to the
simulation results, as BW and D increased, and for = 0.5 and 0.9, the SMT algorithm delivered a better

throughput and PDR performance than the SPT and MST algorithms. At different network parameters,
the CRN with the SMT algorithm offered a better performance than the MST algorithm because it used an
addition node (Steiner node). The use of the addition note shortened the distance path for the transmission of
data from the CR source to the destinations. This enhanced the CRN performance by reducing the time
required for transmission. For a high traffic load of PUs ( 1), and at different network parameters,

the CRN with the SPT algorithm displayed a better performance than the CRN with the SMT or MST
algorithm. This applied for both throughput and PDR performances. It should be noted that as the value of

is increased; the performance of the CRN at the three algorithm protocols is enhanced in terms of throughput
and PDR. This is because at a high value of , the probability that suitable channels (with a low traffic load

of PU) will be available for transmission by CR users, is raised.

ACKNOWLEDGEMENTS
The authors would like to thank Mustansiriyah University (www.uomustansiriyah.edu.iq) Baghdad-
Iraq for its support in the present work.

REFERENCES
[1] Y.C. Liang, Y. Zeng, E. C.Y. Peh and A. T. Hoang, “ Sensing-Throughput Tradeoff for Cognitive Radio
Networks”, IEEE Transactions On Wireless Communications, vol. 7, no. 4, pp. 1326- 1337,APRIL 2008.
[2] M. K. Kaushik, Y. Yoganandam, S. K. Sahoo, “ Sensing and Sharing Schemes for Spectral Efficiency of Cognitive
Radios”, International Journal of Electrical and Computer Engineering (IJECE), vol. 8, no. 5, pp. 2934-2941,
October 2018.
[3] Sh. D. Borde, K. R. Joshi, “Enhanced signal detection algorithm using trained neural network for cognitive radio
receiver ”, International Journal of Electrical and Computer Engineering (IJECE),vol. 9, no. 1,pp. 323-331,
February 2019.
[4] I. F. Akyildiz, W.Y. Lee, M. C. Vuran and Shantidev Mohanty, “NeXt generation/dynamic spectrum
access/cognitive radio wireless networks: A survey”, Computer Networks, vol. 50, no. 13, pp. 2127-2159,
May 2006.
[5] V. Balajia, P. Kabra, P. V. P. K. Saieesh, C. Hota and G. Raghurama, “Cooperative Spectrum Sensing in Cognitive
Radios using Perceptron Learning for IEEE 802.22 WRAN”, Eleventh International Multi-Conference on
Information Processing-2015 (IMCIP-2015), Procedia Computer Science, vol. 54, pp. 14 – 23,2015.
[6] H. Al-Mahdi and Y. Fouad, “Design and analysis of routing protocol for cognitive radio ad hoc networks in
heterogeneous environment”, International Journal of Electrical and Computer Engineering (IJECE), vol. 9, no. 1,
pp. 341-351, February 2019.
Indonesian J Elec Eng & Comp Sci, Vol. 17, No. 1, January 2020 : 291 - 302